Computer system designs that seek to increase computational power while reducing the size of computer equipment create many challenges with controlling the temperature within these ‘dense’ computer systems. Increasing the computational power of computer systems often results in the utilization of high power components that generate high levels of heat. Reducing the size of the computer system often involves packaging components in close proximity to each other, therefore restricting airflow through the system. The combination of high power, high heat generating components and compact design is pushing the limits of current air-cooled systems.
Air-cooled systems often utilize an array of fans to move air from the environment, through a computer chassis, and back to the environment. As the air passes through the enclosure it comes in thermal contact with, and absorbs heat from, the heat generating components within the enclosure. The heat transfer rate that can be achieved by an air-cooled system is a function of the volumetric flow rate and static pressure of air that can be moved through the chassis.
The performance of many air-cooled systems is also dependent on the temperature of air available to the system. As the temperature of air drawn into an air-cooled system increases, the amount of heat that the air can absorb decreases, thus decreasing the effectiveness of the cooling system. This is often of particular concern in data centers and other installations where large concentrations of dense computer systems operate.